forecasting climate change impacts on large pelagic fish ... · forecasting climate change impacts...

Forecasting climate change impacts on large pelagic fish

populations and fisheries Patrick Lehodey1, Inna Senina1, Simon Nicol2, John Hampton2, Anna Conchon1, Anne-Cecile Dragon1, Olivier Titaud1, Beatriz Calmettes1, Olivier Aumont3, Morgane Dessert3, Thomas Gorgues4, Matthieu Lengaigne4, Christophe Menkes4

1CLS, Space Oceanography Division, France. E-mail: [email protected] 2 SPC, Oceanic Fisheries Programme, New Caledonia. 3 IRD, LPO, IUEM France 4 IRD, LOCEAN, New Caledonia

3rd PICES/ICES/IOC Symposium on “Effects of Climate Change on the World’s Oceans” Mar 23-27, 2015, Santos City, Brazil

Outline of the presentation Overview of High Seas Fisheries

Key variables and processes Modeling Projecting fish population dynamics Perspectives

Albacore

Yellowfin Skipjack

Bigeye

Swordfish

Tunas Billfishes

Marlins

Sailfishes bluefin

By-catches and protected (iconic) species

High Sea Fisheries

High Sea Fisheries

Longline – yellowfin, bigeye, albacore, swordfish, bluefin

Purse seine – skipjack, yellowfin, small amount of bigeye

Friendly from Alain Fonteneau

Mean annual catch over 1992-2011

Source: FAO

155,557,000 km2 (52%)

Pacific

76,762,000 km2 (25%)

Atlantic

Area (60°N/S): 68,556,000 km2 (23%)

Indian Th

ousa

nds

tonn

es

High Sea Fisheries

Mean Chl_a

(mg m3) 1998-2012

Mean SST (°C) 1980-2000

High Sea Fisheries

Friendly from Alain Fonteneau

Mean Chl_a (mg m3) 1998-2012

How can CC impact distributions, abundances, catch in the FUTURE?

Can we predict the PAST and explain such differences?

Key environmental variables and mechanisms

Species Fork length (cm)

Lower lethal O2 levels (ml l-1)

Skipjack 50 1.87 Albacore 50 1.23 Yellowfin 50 1.14 Bigeye 50 0.40

Estimated lower lethal oxygen

Species

Temperature (°C)

Skipjack 20-29 Yellowfin 20-30 Bigeye 13-27 Albacore 15-21 Sth. bluefin 17-20

Range of sea surface temperature with substantial catches

Graham 1973; Sharp & Dizon 1977; Sund et al. 1981; Brill et al 1987

Key environmental variables

Tem

pera

ture

Diss. oxygen

Skipjack

Albacore

Bigeye

Yellowfin

Temperature & oxgen

Distribution of skipjack catch in the Pacific O. and SST (°C). a) First half of 1989 (La Niña phase) and b) first half of 1992 (El Niño phase ).

Lehodey et al. (1997)

Displacement of tagged skipjack, a: released in Apr 1991 and recaptured before Feb 1992 (El Niño phase) ; b, released in May 1991 and recaptured before Feb 1992 (El Niño phase); c: released in Mar 1992 and recaptured before Oct 1992 (La Niña phase).


El Niño

La Niña

Temperature & oxgen

Population basin-scale redistribution of skipjack population with ENSO warm and cold phases

El Niño

El Niño

La Niña

50%

Dufour et al. (2010)

Key environmental variables Temperature & oxgen

Bluefin tuna cumulative catch by fisheries in the Bay of Biscay in relation to Julian date

Julia

n da

y 5%

50%

95%

Bluefin tuna arrival occurred 14 days earlier in the last 5 years than in the first 5 years of the time series (which represents a rate of change of 5.6 days per decade).

Phenological change in feeding migration


Bluefin tuna were captured east of Greenland (65°N) during exploratory fishing in Aug 2012 together with 6 tonnes of mackerel, which is a preferred prey species and itself a new immigrant to the area. “The presence of bluefin in this region is likely due to a combination of warm temperatures [increasing since 1985] and immigration of an important prey species to the region.” MacKenzie et al (2014)

Temperature & oxgen

Key environmental variables Vertical tracking of blue marlin vs dissolved oxygen concentration in: western tropical Atlantic Eastern tropical Atlantic

Temperature & oxgen

Time-series since 1960 of dissolved O2 near 170°W at the equator (5°S-5°N) showing expansion of Oxygen Mimum Zone. (Stramma et al. 2008)

Stramma et al (2011)

Population distribution results from individual behaviors in a 3D environment Behavior is constrained by physiological (T° &O2) aptitudes (species and age dependent)


Time series of depth and temperature for one bigeye tuna tagged in the N-W Atlantic (C. H. Lam et al. 2014)

night day

Prey detection from acoustic 38 kHz

Lunch

Breakfast Dinner

thermoregulation

Within these limits basic requirements are :

• Feeding • Reproduction • Escaping predator

Mean Chl_a (mg m3) 1998-2012 Larvae

Tuna (yellowfin) daily food ration = 4-7 % of their body weight (Olson et al 1986)

Crédit: Rudy Kloser et Jock Young CSIRO, Australie

Micronekton

The forage of oceanic predators

Key environmental variables Food

Modeling

Approach Questions Link to CC

Ecological models Habitat models Environmental Niche (~Bioclimatic envelop) models Trophic network

• Identification of key variables • Species distributions • Favorability (index)

Coupling to CC models • Spatial shifts in distribution? • Relative changes in favorability? • Local species richness turnover? • Global productivity indicators

Standard Pop. Dynamics Models • Virtual Population Analysis (e.g. ADAPT: Least square; no error in catch) • Age-Structure population models with multiples data sources /areas (Stock synthesis; Multifan-CL: MLE with Bayesian approach for error process)

• Short term Management • Abundance estimate, • Fishing impact, • Sustainability, indicators • Fishing scenarios

No coupling to CC models • No projection (or equilibrium)

Hybrid models Combining target population dynamics with 2 or 3D Environmental variables or « Ecosystem » models driving key mechanisms

• Long term (typically multi-decadal) trends • Spatial Management and fishing scenarios

Coupling to CC models is CC adding negative or positive stress to the stocks? • When, where? • How to adapt management and fishing mortality?

Modeling Toolbox: Models vs Questions

Lehodey et al. (1998, 2010, 2014) The SEAPODYM model includes 6 functional groups of micronekton with spatial and temporal dynamics based on time of development, linked to temperature.

Jennings et al. (2008) PP and temperature data are used to predict the biomass and production of marine animals on a global scale based on relationships between body size, energy acquisition and transfer.

Teleost biomass and FAO fishing areas Micronekton (prey of large pelagics)

Maury et al (2007); Lefort et al (2015) The APECOSM model is a size-structured bio-energetic model that simulates 3D dynamical distributions of three interactive pelagic communities (epipelagic, mesopelagic, and migratory)

mol C m-2 Mesopelagic community

Total biomass (1985-05) of 6 micronekton groups

Modeling

Closing the life cycle

Feeding habitat = Abundance of prey

(micronekton) x accessibility (T°,O2)

Spawning habitat = Temperature Prey (zoopl.)

Predators (micronekton)

IF MATURE seasonal switch

Movement to

Spawning areas

(currents)

Mortality

Spawning success

Recruitment

Movement to

feeding areas (currents)

+++ +

-

By cohort

Age structured population

1 yr 2 yr

…

Size

+

Nb

indi

vidu

als

Modeling Hybrid model Spatial Ecosystem And

Population Dynamics Model SEAPODYM

Growth mortality

Fishing

Lehodey et al. (2003,2008) Senina et al (2008) www.spc.int/ofp/seapodym/

Model parameter Estimation (MLE)

(catch, size, acoustic and tagging data)

T, O2, PP, currents

Test same parameterization

in another Basin

1st Phase: Rebuilding the past history and testing the model parameterization

Total predicted skipjack density and observed catch (% to cicrcles)

Modeling Hybrid model SEAPODYM

Tropical skipjack vs temperate albacore

Modeling Hybrid model SEAPODYM

Projections

Projections SEAPODYM tuna projections

Model optimization with historical fishing

(1960-2010)

Projection

Initial Conditions Parameters

Historical fishing (WCPFC,

IATTC, ICCAT, IOTC, …)

Fishing effort scenarios

Coupled ocean-biogeochemical

models forced by atmospheric

reanalyses (NCEP, ERA-INTERIM)

Earth Climate Models (Atmospheres-Ocean-Land-Ice) coupled to

biogeochemical model(s) forced by

IPCC greenhouse gas release scenarios

Problem: Earth Climate Model have biases

Environmental forcing Fishing forcing

Projections Anomaly between models

and Word Ocean Atlas climatology

Mean sea surface

temperature for historical period

Forced from atmospheric reanalysis (NEMO-INTERIM)

Earth Climate model (IPSL – CM5)

Wexler et al. (2011)

Optimal range for growth of Yellowfin tuna larvae is 26-31oC with low and high lethal temperatures of 21 and 33oC respectively.

Biases can have strong consequences, especially using absolute rather than relative parameterization

SKIPJACK LARVAE (A2 scenario)

2000

20

50

2099

1st Exp with IPSL-CM4 2nd Exp after T° correction

≠ 4°C

Temperature transect at longitude 180°

Bias correction

Lehodey et al.(2013)

Projections

2000s

2090s

Larvae Pacific Skipjack LarvaeSouth Pacific albacore

Lehodey et al. 2013, 2014

2090s

2000s

Projections SP Alb. pop. biomass

Perspectives

Perspectives Micronekton

CC Forcing

Comparisons Missing

mechanisms? Management

Atmospheric CO2

Light, nutrients Phytoplankton

photosynthesis

Detritus

Zooplankton

DV

M

Zooplankton

Passive Flux Zooplankton Flux

Sediment trap

Mix

ing,

upw

ellin

g Microbial loop

Sea surface

Uncertainty on the global biomass estimates at least 1 order of magnitude!

~31% of anthropogenic CO2 absorbed by the Ocean The amount of C exported from surface layer (“biological pump”) and estimated from direct measures of passive transport (traps) is lower than estimations by other methods (models), sometimes by 70% !

Migrant (day) micronekton

Non migrant micronekton

Migrant (night) micronekton

DV

M

Micronekton Mediated Export

Davison et al (2013) estimated that migrant micronekton participate between <10% (mesotrophic) and >40 % (oligotrophic) of total carbon export in the Calcofi region!


CC Forcing

Comparisons

Missing


For Biological/Ecological modelling we need ensemble simulation projections with the same mean state and realistic historical conditions Easy access to the research community after validation of datasets

2m-temperature (°C) averaged on the El Nino 3.4 Box for the historical (black) and future periods (red for IPSL, green for GFDL and blue for NorESM).

Historical reanalysis


CC Forcing

Comparisons

Missing


Models inter-comparison require to use the same environmental AND fishing forcing. The use of historical fishing data require some expertise (better to publish methods and datasets) Existing initiative (ISI-MIP)

Dueri et al (2014, apecosm-e)

Skipjack adult biomass (1990–99)

Biomass by ocean

Lehodey et al (2013, seapodym) Jennings et al., (2008) Cheung et al . (2010)

Historical fishing dataset published in Earth System

Sciences Data (Lehodey et al 2014)


CC Forcing

Comparisons

Missing


What could be the impact of ocean acidifiation? Species adaptation (genetic)? Competition?

Prey of larvae

Predator of larvae

°C

Atl. Bluefin Spawning index

Λ

Spawning habitat mechanisms

24 Juin 2002

mean pCO2 and larval survival after 7 days (Bromhead et al 2014)


CC Forcing

Comparisons Missing


Trends in O2 concentration and temperature stratification will (very likely) increase catchability by purse seiners of skj & yft (& juvenile bigeye) in the surface layer of several oceanic regions.

Dissolved O2 near 170°W at the equator (Stramma et al. 2008)

Increasing access to tuna stocks in international waters Need smart management mechanisms accounting for new monitoring technologies!

Projection of skipjack total biomass A2 scenario. (Bell et al. 2013)

2005

2035 IATTC

WCPFC

PNA

Sibert et al 2012

http://www.imber.info/index.php/Science/Regional-Programmes/CLIOTOP

Thanks to PICES for the support and invitation!

Deadline for abstract submission 31 March !

forecasting climate change impacts on large pelagic fish ... · forecasting climate change impacts...

Documents